Integrative Transcriptomic Meta-Analysis Reveals Risk Signatures and Immune Infiltration Patterns in High-Grade Serous Ovarian Cancer

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Summary: High-grade serous ovarian cancer (HGSOC) is an aggressive malignancy with poor clinical outcomes due to late-stage diagnosis and the development of chemoresistance. The 5-year survival rate remains below 30%. To uncover underlying mechanisms and identify clinically relevant targets, the authors performed a gene expression meta-analysis using publicly available microarray datasets. By analyzing data from 11 studies (291 HGSOC and 96 control samples), they identified 892 differentially expressed genes (DEGs) significantly associated with HGSOC.

Methodically, the authors started with global transcriptomic profiling, moved into interaction networks, assessed clinical relevance, explored immune context, and finally concluded with single-cell transcriptomic resolution. Key pathways were identified, including mitochondrial dysfunction, vesicle trafficking, ECM remodeling, and immune regulation—all known contributors to HGSOC pathogenesis.

Overall, it’s a comprehensive and well-structured study that connects molecular insights with clinical relevance. This study provides translational value and paves the way for future work in cancer precision medicine.

A few points would need clarification:

• Fig 3 (PPI): Which clusters or genes function as central hubs in the network? Are there ways that can help us prioritize targets?

• Single-cell analysis (Fig 7): CAST and TGOLN2 show broad expression, but does that dilute their value as therapeutic targets?

• Is it possible to examine whether the gene expression patterns observed within specific cell clusters correspond to distinct functional phenotypes? For example, the cell populations expressing PTGIS and TGFBR2 exhibit proliferative or quiescent characteristics?

Author Response

Please find attached the document containing all the revisions made to our manuscript

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript aims to establish an immune subtype-based classification system using transcriptomic data, with the goal of identifying prognostic immune-related signatures and predicting immunotherapy responses. The study is timely and contributes to the rapidly growing field of tumor immune landscape analysis. However, there are several points that require clarification and improvement before the manuscript is suitable for publication. 
1.    Please clarify how immune subtypes were defined and whether clustering robustness was evaluated. The immune classification forms the foundation of this work, but the clustering method (e.g., k-means, consensus clustering, NMF) and the basis for choosing the number of clusters (k=?) are not clearly described. The authors should provide the detailed method, including how the cluster number was determined (e.g., using the cumulative distribution function or? silhouette score), and whether the stability and reproducibility of clusters were validated.
2.    The authors should compare their immune subtypes with known classifications in the literature. Are the current subtypes overlapping with any existing systems?  This comparison will help readers understand the novelty of the proposed model.
3.    Please improve the methodology description of immune infiltration estimation. It is unclear which algorithm was used to estimate immune cell proportions. Was it CIBERSORT, xCell, ssGSEA, or EPIC? Also, were the RNA-seq data normalized appropriately for deconvolution? If multiple algorithms were applied, please explain how the consistency among them was handled.
4.    The prognostic significance of the immune clusters should be validated in external datasets. Currently, the survival analysis is based only on the TCGA cohort. For the results to be more convincing, the authors are encouraged to perform validation using an independent GEO dataset, especially for the prognostic relevance of immune subtypes or signatures.
5.    Please strengthen the interpretation of immune checkpoint expression and ICB response prediction. Although the manuscript discusses expression levels of PD-1/PD-L1, CTLA-4, etc., the authors should include more detailed discussion or figures showing TIDE or Immunophenoscore results, and clarify whether any subtype shows predictive advantage. More importantly, real-world clinical cohorts should be referenced, if available, to enhance translational significance.
6.    Figures are too low in resolution; please provide high-resolution versions with clearer legends. In particular, Figures 2 and 3 have very small font sizes and overlapping annotations. Uniform color usage and enlarging key labels would greatly help readers.

Author Response

Please find attached the document containing all the revisions made to our manuscript

Author Response File: Author Response.pdf